Figure 7
Robert Froemel & Lars Velten
2025-03-12
Figure 7A
As in figure 2, determine which combinations result in more or less activity, compared to random DNA controls.
heptad <- unique(c(mpra.data$HSPC.libB$DATA$TF1.name,unique(mpra.data$HSPC.libB$DATA$TF2.name)))
final.crs.filtered <- mpra.data$HSPC.libC.aggregate$DATA
all_bad_sites <- subset(mpra.data$HSPC.libC.aggregate$CONTROLS.GENERAL, ControlType %in% c("ScrambleTFBS"))
bad_range <- ddply(as.data.frame(all_bad_sites), "clusterID",summarise, q05.background = quantile(mean.norm.adj, 0.05), q95.background = quantile(mean.norm.adj,0.95))
rownames(bad_range) <- paste(bad_range$clusterID, bad_range$Library, sep="_")
final.crs.filtered$tf1 <- with(final.crs.filtered, ifelse(TF1.name < TF2.name, TF1.name, TF2.name))
final.crs.filtered$tf2 <- with(final.crs.filtered, ifelse(TF1.name < TF2.name, TF2.name, TF1.name))
final.crs.filtered$tfcombo <- paste(final.crs.filtered$tf1, final.crs.filtered$tf2, sep="-")
final.crs.filtered$nrepeats <- final.crs.filtered$TFnumber
final.crs.filtered <- subset(final.crs.filtered,select = c("mean.norm.adj","tfcombo", "clusterID"))
active_by_celltype_sum <- ddply(as.data.frame(final.crs.filtered), c("tfcombo"), summarise,
n_obs_bylib = length(mean.norm.adj),
n_active_bylib = sum(mean.norm.adj > bad_range[unique(clusterID),"q95.background"]),
n_repressed_bylib = sum(mean.norm.adj < bad_range[unique(clusterID),"q05.background"]))
active_m_separate <- ddply(active_by_celltype_sum, c("tfcombo"),summarise,
n_obs_total = sum(n_obs_bylib),
max_active = max(n_active_bylib/n_obs_bylib), max_repressed = max(n_repressed_bylib/n_obs_bylib),
total_active = sum(n_active_bylib)/n_obs_total,
total_repressed = sum(n_repressed_bylib)/n_obs_total)
active_by_celltype <- ddply(active_by_celltype_sum, c("tfcombo"), summarise,
n_obs = sum(n_obs_bylib),
n_active = sum(n_active_bylib) / n_obs,
n_repressed = sum(n_repressed_bylib) / n_obs)
active_m <- ddply(active_by_celltype, "tfcombo",summarise,
n_obs_total = sum(n_obs),
total_active = mean(n_active), total_repressed = mean(n_repressed),
max_active = max(n_active), max_repressed = max(n_repressed),
var_active = sd(n_active), var_repressed = sd(n_repressed), #for the vatiance need resampling based w random DNA
ct_spec_act = max_active - min(n_active), ct_spec_repr = max_repressed -min(n_repressed))
### Establish background rage ####
n_simulated_tf <- 1000 #arbitrary
n_seq_per_tf <- round(mean(table(final.crs.filtered$tfcombo)))
all_bad_sites_libB <- all_bad_sites
resampled_bad_sites_by_celltype_libB <- all_bad_sites_libB[sample(1:nrow(all_bad_sites_libB), n_simulated_tf * n_seq_per_tf,replace = T),]
resampled_bad_sites_by_celltype_libB$tfcombo <- rep(1:n_simulated_tf, each = n_seq_per_tf)
resampled_bad_sites_by_celltype_libB <- ddply(as.data.frame(resampled_bad_sites_by_celltype_libB), c("tfcombo","clusterID"), summarise,
n_active = mean(mean.norm.adj > bad_range[unique(clusterID),"q95.background"]),
n_repressed = mean(mean.norm.adj < bad_range[unique(clusterID),"q05.background"]))
bad_sites_summary_libB <- ddply(resampled_bad_sites_by_celltype_libB, "tfcombo",summarise,
total_active = mean(n_active), total_repressed = mean(n_repressed),
max_active = max(n_active), max_repressed = max(n_repressed),
var_active = sd(n_active), var_repressed = sd(n_repressed))
lines_act_libB <- quantile(bad_sites_summary_libB$max_active, c(0.025,0.975))
lines_rep_libB <- quantile(bad_sites_summary_libB$max_repressed, c(0.025,0.975))
lines_libB <- list(lines_act_libB,lines_rep_libB)
active_m_libB <- active_m_separate
activators_libB <- active_m_libB$tfcombo[active_m_libB$max_active > lines_act_libB[2]]
represssors_libB <- active_m_libB$tfcombo[active_m_libB$max_repressed > lines_rep_libB[2]]
active_m$activator <- active_m$tfcombo %in% activators_libB
active_m$repressor <- active_m$tfcombo %in% represssors_libB
active_m$class <- with(active_m, ifelse(activator & repressor, "dual",
ifelse(activator, "activator",
ifelse(repressor, "repressor","other"))))
active_m$TF1 <- gsub(".+-","",active_m_libB$tfcombo)
active_m$TF2 <- gsub("-.+","",active_m_libB$tfcombo)
flipped <- active_m
flipped$TF1 <- active_m$TF2
flipped$TF2 <- active_m$TF1
combined <- rbind(active_m, flipped)
saveRDS(active_m, file = "Figure7_TFpairs_HSPC.rds")
active_libA <- readRDS("~/cluster/project/SCG4SYN/Manuscript/1_libA_levels_HSC/003_activem_libA_updated.rds")
# new_classes <- subset(readRDS("/users/lvelten/project/SCG4SYN/Manuscript/2_libH/003_revised_classification.rds"), !grepl("pairs", faceting), select = c("tfcombo", "class"))
# colnames(new_classes)[1] <- "TF"
# active_libA <- merge(active_libA[,-13], new_classes)
# active_libA$class <- factor(active_libA$class, levels = rev(c("activator","dual","other","repressor")))
# saveRDS(active_libA, file = "/users/lvelten/project/SCG4SYN/Manuscript/1_libA_levels_HSC/003_activem_libA_updated.rds")
levels.tf1 <- readRDS("additional_data/Figure7_TForder.rds")
combined$TF1 <- factor(combined$TF1, levels = levels.tf1)
combined$TF2 <- factor(combined$TF2, levels = levels.tf1)
saveRDS(combined, file = "Figure7_TFpairs_HSPC_full.rds")
a <- ggplot(aes(x = TF1, y = TF2),data = subset(combined, as.integer(TF1)<as.integer(TF2))) +
geom_tile(aes(fill = 100* max_active)) +
theme_bw()+ theme(axis.text.x = element_text(angle=90),panel.grid = element_blank(), axis.title = element_blank()) +
scale_x_discrete(limits = levels(combined$TF1),position = "top") + scale_y_discrete(limits = levels(combined$TF1))+
scale_fill_gradientn(colours = c("black", "red", "orange", "yellow"), limits = c(0,100), name = "% active") +
new_scale_fill() +
geom_tile(aes(fill = 100* max_repressed), data = subset(combined, as.integer(TF1)>as.integer(TF2))) +
scale_fill_gradientn(colours = c("black", "blue", "green"), limits = c(0,30), name = "% repressed") +
geom_point(shape=16, color = "white", size = 0.5, data = subset(combined, as.integer(TF1)<as.integer(TF2) & class %in% c("activator", "dual"))) +
geom_point(shape=16, color = "white", size = 0.5, data = subset(combined, as.integer(TF1)>as.integer(TF2) & class %in% c("repressor", "dual"))) +
scale_shape_manual(values = c("activator" = 2, "repressor" = 6, "dual" = 3)) #+
#scale_color_manual(values = c("activator" = "#65C1A4", "dual"="#F68D62", "other"="#8C9DC9", "repressor"="#E28BBB"), name = "Class (combination)")+
#geom_point(data =subset(combined,TF1.class == "activator" & TF2.class == "activator" & class %in% c("repressor","dual") ), color = "white", size = 3, shape =4)
print(a)## Warning: No shared levels found between `names(values)` of the manual scale and the
## data's shape values.
Figure 7B
final.crs.filtered <- mpra.data$K562.libC.minP.tra$DATA
all_bad_sites <- subset( mpra.data$K562.libC.minP.tra$CONTROLS.GENERAL, ControlType %in% c("ScrambleTFBS"))
bad_range <- ddply(as.data.frame(all_bad_sites), c("clusterID", "Library"),summarise, q05.background = quantile(mean.norm.adj, 0.05), q95.background = quantile(mean.norm.adj,0.95))
rownames(bad_range) <- paste(bad_range$clusterID, bad_range$Library, sep="_")
final.crs.filtered$tf1 <- with(final.crs.filtered, ifelse(TF1.name < TF2.name, TF1.name, TF2.name))
final.crs.filtered$tf2 <- with(final.crs.filtered, ifelse(TF1.name < TF2.name, TF2.name, TF1.name))
final.crs.filtered$tfcombo <- paste(final.crs.filtered$tf1, final.crs.filtered$tf2, sep="-")
final.crs.filtered$nrepeats <- final.crs.filtered$TFnumber
final.crs.filtered <- subset(final.crs.filtered,select = c("mean.norm.adj","tfcombo", "clusterID", "Library"))
# toadd <- subset(libH.tp53, select = c("mean.norm.adj","clusterID"), nrepeats <= 3 & affinitynum >= 0.5)
# toadd$tfcombo <- "Tp53 only"
# final.crs.filtered <- rbind(final.crs.filtered, toadd)
active_by_celltype_sum <- ddply(as.data.frame(final.crs.filtered), c("tfcombo","clusterID", "Library"), summarise,
n_obs_bylib = length(mean.norm.adj),
n_active_bylib = sum(mean.norm.adj > bad_range[paste(unique(clusterID), unique(Library), sep="_"),"q95.background"]),
n_repressed_bylib = sum(mean.norm.adj < bad_range[paste(unique(clusterID), unique(Library), sep="_"),"q05.background"]))
active_m_separate <- ddply(active_by_celltype_sum, c("tfcombo","Library"),summarise,
n_obs_total = sum(n_obs_bylib),
max_active = max(n_active_bylib/n_obs_bylib), max_repressed = max(n_repressed_bylib/n_obs_bylib),
total_active = sum(n_active_bylib)/n_obs_total,
total_repressed = sum(n_repressed_bylib)/n_obs_total)
active_by_celltype <- ddply(active_by_celltype_sum, c("tfcombo", "clusterID"), summarise,
n_obs = sum(n_obs_bylib),
n_active = sum(n_active_bylib) / n_obs,
n_repressed = sum(n_repressed_bylib) / n_obs)
active_m <- ddply(active_by_celltype, "tfcombo",summarise,
n_obs_total = sum(n_obs),
total_active = mean(n_active), total_repressed = mean(n_repressed),
max_active = max(n_active), max_repressed = max(n_repressed),
var_active = sd(n_active), var_repressed = sd(n_repressed), #for the vatiance need resampling based w random DNA
ct_spec_act = max_active - min(n_active), ct_spec_repr = max_repressed -min(n_repressed))
n_simulated_tf <- 1000 #arbitrary
n_seq_per_tf <- round(mean(table(final.crs.filtered$tfcombo)))
all_bad_sites_libB <- subset(all_bad_sites, Library == "LibB")
resampled_bad_sites_by_celltype_libB <- all_bad_sites_libB[sample(1:nrow(all_bad_sites_libB), n_simulated_tf * n_seq_per_tf,replace = T),]
resampled_bad_sites_by_celltype_libB$tfcombo <- rep(1:n_simulated_tf, each = n_seq_per_tf)
resampled_bad_sites_by_celltype_libB <- ddply(as.data.frame(resampled_bad_sites_by_celltype_libB), c("tfcombo","clusterID"), summarise,
n_active = mean(mean.norm.adj > bad_range[paste(unique(clusterID), unique(Library), sep="_"),"q95.background"]),
n_repressed = mean(mean.norm.adj < bad_range[paste(unique(clusterID), unique(Library), sep="_"),"q05.background"]))
bad_sites_summary_libB <- ddply(resampled_bad_sites_by_celltype_libB, "tfcombo",summarise,
total_active = mean(n_active), total_repressed = mean(n_repressed),
max_active = max(n_active), max_repressed = max(n_repressed),
var_active = sd(n_active), var_repressed = sd(n_repressed))
lines_act_libB <- quantile(bad_sites_summary_libB$max_active, c(0.025,0.975))
lines_rep_libB <- quantile(bad_sites_summary_libB$max_repressed, c(0.025,0.975))
lines_libB <- list(lines_act_libB,lines_rep_libB)
active_m_libB <- subset(active_m_separate, Library == "LibB")
activators_libB <- active_m_libB$tfcombo[active_m_libB$max_active > lines_act_libB[2]]
represssors_libB <- active_m_libB$tfcombo[active_m_libB$max_repressed > lines_rep_libB[2]]
active_m$activator <-active_m$tfcombo %in% activators_libB
active_m$repressor <-active_m$tfcombo %in% represssors_libB
active_m$class <- with(active_m, ifelse(activator & repressor, "dual",
ifelse(activator, "activator",
ifelse(repressor, "repressor","other"))))
saveRDS(active_m, file = "Figure7_TFpairs_K562.rds")
active_m$TF1 <- gsub(".+-","",active_m_libB$tfcombo)
active_m$TF2 <- gsub("-.+","",active_m_libB$tfcombo)
flipped <- active_m
flipped$TF1 <- active_m$TF2
flipped$TF2 <- active_m$TF1
combined <- rbind(active_m, flipped)
combined$TF1 <- factor(combined$TF1, levels = levels.tf1)
combined$TF2 <- factor(combined$TF2, levels = levels.tf1)
saveRDS(combined, file = "Figure7_TFpairs_K562_full.rds")
a<- ggplot(aes(x = TF1, y = TF2),data = subset(combined, as.integer(TF1)<as.integer(TF2))) +
geom_tile(aes(fill = 100* max_active)) +
theme_bw()+ theme(axis.text.x = element_text(angle=90),panel.grid = element_blank(), axis.title = element_blank()) +
scale_x_discrete(limits = levels(combined$TF1),position = "top") + scale_y_discrete(limits = levels(combined$TF1))+
scale_fill_gradientn(colours = c("black", "red", "orange", "yellow"), limits = c(0,100), name = "% active") +
new_scale_fill() +
geom_tile(aes(fill = 100* max_repressed), data = subset(combined, as.integer(TF1)>as.integer(TF2))) +
scale_fill_gradientn(colours = c("black", "blue", "green"), limits = c(0,30), name = "% repressed") +
geom_point(shape=16, color = "white", size = 0.5, data = subset(combined, as.integer(TF1)<as.integer(TF2) & class %in% c("activator", "dual"))) +
geom_point(shape=16, color = "white", size = 0.5, data = subset(combined, as.integer(TF1)>as.integer(TF2) & class %in% c("repressor", "dual"))) +
scale_shape_manual(values = c("activator" = 2, "repressor" = 6, "dual" = 3))
print(a)## Warning: No shared levels found between `names(values)` of the manual scale and the
## data's shape values.
Figure 7C
combined_k562 <- readRDS("Figure7_TFpairs_K562_full.rds")
combined_k562$class <- factor(combined_k562$class, levels = c("activator","dual","other","repressor"))
combined <- readRDS("Figure7_TFpairs_HSPC_full.rds")
combined$class <- factor(combined$class, levels = c("activator","dual","other","repressor"))
#add annotation on the single factors
active_libA <- subset(readRDS("Figure2_TF_classification.rds"), !grepl("-", tfcombo))
classes.libA <- active_libA$class
names(classes.libA) <- active_libA$tfcombo
combined$TF1.class <- classes.libA[as.character(combined$TF1)]
combined$TF2.class <- classes.libA[as.character(combined$TF2)]
libA_K562 <- readRDS("additional_data/Figure7_K562_singleFactorClass.rds")
classes.libA <- libA_K562$class
names(classes.libA) <- libA_K562$TF
combined_k562$TF1.class <- classes.libA[as.character(combined$TF1)]
combined_k562$TF2.class <- classes.libA[as.character(combined$TF2)]
stats.k562 <- with(subset(combined_k562, as.integer(TF1)<as.integer(TF2) ), table(class))
stats.hspc <- with(subset(combined, as.integer(TF1)<as.integer(TF2) ), table(class))
switches.k562<- with(subset(combined_k562, as.integer(TF1)<as.integer(TF2) & (TF1.class == "activator" & TF2.class == "activator" & class %in% c("repressor","dual"))), length(class))
switches.hspc <- with(subset(combined, as.integer(TF1)<as.integer(TF2) & (TF1.class == "activator" & TF2.class == "activator" & class %in% c("repressor","dual"))), length(class))
stats <- data.frame(class = c(names(stats.k562), "switches", names(stats.hspc), "switches"),
n = c(unname(stats.k562),switches.k562, unname(stats.hspc), switches.hspc),
celltype = rep(c("K562", "HSPC"), each = 5))
stats$class[stats$class == "other"] <- "inactive"
stats$class <- factor(stats$class, levels = c("activator", "other", "repressor", "dual"))
a <- qplot(x = celltype, y = n, fill = class, data = subset(stats, class %in% c("activator", "repressor")), geom="col") + facet_wrap(~class, scales= "free_y", ncol=1) + theme_bw() +
theme(panel.grid = element_blank()) + ylab("Number of factor pairs") + xlab("Cell type") + scale_fill_manual(values = c("activator" = "#65C1A4", "dual"="#F68D62", "other"="#8C9DC9", "repressor"="#E28BBB"), name = "Class (combination)")## Warning: `qplot()` was deprecated in ggplot2 3.4.0.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.print(a)
Figure 7D
stats.ar <- subset(combined, as.integer(TF1)<as.integer(TF2) & ((TF1.class == "activator" & TF2.class == "repressor") |(TF2.class == "activator" & TF1.class == "repressor") ), select = c("tfcombo", "class", "TF1", "TF2"))
stats.ar$celltype = "HSPC"
stats.ar.k562 <- subset(combined_k562, as.integer(TF1)<as.integer(TF2) & ((TF1.class == "activator" & TF2.class == "repressor") |(TF2.class == "activator" & TF1.class == "repressor") ), select = c("tfcombo", "class", "TF1", "TF2"))
stats.ar.k562$celltype = "K562"
stats.ar <- rbind(stats.ar, stats.ar.k562)
stats.ar$class <- factor(stats.ar$class, levels = c("activator", "other", "repressor", "dual"))
stats.ar.norm <- ddply(stats.ar, "celltype", summarise, variable = levels(class), freq = table(class) / length(class))## Warning: Returning more (or less) than 1 row per `summarise()` group was deprecated in
## dplyr 1.1.0.
## ℹ Please use `reframe()` instead.
## ℹ When switching from `summarise()` to `reframe()`, remember that `reframe()`
## always returns an ungrouped data frame and adjust accordingly.
## ℹ The deprecated feature was likely used in the plyr package.
## Please report the issue at <https://github.com/hadley/plyr/issues>.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.a <- qplot(x = celltype, y = 100*freq, fill = variable, geom="col",data = stats.ar.norm) + theme_bw() + scale_fill_brewer(palette = "Set2", name = "Class (Combination)") +
theme(panel.grid = element_blank()) + ylab("% of activator-repressor pairs") + xlab("Cell type")
print(a)## Don't know how to automatically pick scale for object of type <table>.
## Defaulting to continuous.
Figure 7E
Combine with activity estimates from Lib A (Figure 2)
active_libA <- readRDS("Figure2_TF_activity.rds")
classes_libA <- subset(readRDS("Figure2_TF_classification.rds"), !grepl("-", tfcombo))
colnames(classes_libA)[1] <- "TF"
active_libA <- merge(active_libA, classes_libA[,c("TF", "class")])
combo_summary <- ddply(combined, "TF1", summarise, nact = sum(class =="activator" | class == "dual"),
nrep = sum(class=="repressor"| class == "dual"), ndual =sum( class == "dual"),
nswitch = sum(TF1.class == "activator" & TF2.class == "activator" & class %in% c("repressor","dual"), na.rm=T))
colnames(combo_summary)[1] <- "TF"
combo_summary <- merge(combo_summary, active_libA)
qplot(x = 100*max_active, y = nact,data = combo_summary, color = class) +
scale_color_manual(values = c("activator" = "#65C1A4", "dual"="#F68D62", "other"="#8C9DC9", "repressor"="#E28BBB"), name = "Class (individual)", guide=F)+
coord_cartesian(xlim=c(-5,100),ylim=c(0,41))+
geom_text_repel(aes(label = TF), data = subset(combo_summary, TF %in% c("Myb", "Trp53", "Cebpa")), color = "black") +
theme_bw() + theme(panel.grid = element_blank()) + scale_y_continuous(breaks = seq(0,40,by=5), labels = sprintf("%d/41",seq(0,40,by=5))) +
xlab("% of sequences active\n(single factor / lib 1)") + ylab("# of combinations activator (lib 3)")## Warning: The `guide` argument in `scale_*()` cannot be `FALSE`. This was deprecated in
## ggplot2 3.3.4.
## ℹ Please use "none" instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
Figure 7F
Fit the relationship between activity of the individual factors and activity of the pair
#calculate an overall slope!!!!!
formerge <- active_libA; colnames(formerge)[1] <- "TF2"
all.merged <- merge(formerge,combined, by = "TF2", suffixes = c(".TF2", ".combined"))
formerge <- active_libA; colnames(formerge)[1] <- "TF1"
all.merged <- merge(all.merged[,c("TF1", "TF2", "max_active.combined", "max_active.TF2")],formerge[,c("TF1", "max_active")], by = "TF1")
all.merged$max_active.combined <- 100*all.merged$max_active.combined
all.merged$max_active <- 100*all.merged$max_active
all.merged$max_active.TF2 <- 100*all.merged$max_active.TF2
all.merged$tfcombo <- paste(all.merged$TF1, all.merged$TF2, sep = "-")
m <- lm( max_active.combined ~ max_active.TF2* max_active + 0, data = all.merged)
grid <- expand.grid(max_active.TF2 = 1:100, max_active = 1:100)
grid$max_active.combined <- predict(m, grid)
gridmat <- acast(max_active ~max_active.TF2, data = grid, value.var = "max_active.combined")
library(plotly)
fir <- plot_ly(z = ~grid) %>% add_surface()
t1 <- list(
size = 16
)
fir <- plot_ly(all.merged, y = ~max_active, x = ~max_active.TF2, z = ~max_active.combined) %>%
add_trace(size = 0.2, color = I("grey30"), text = ~tfcombo, hoverinfo = "info") %>%
add_surface(y = NULL, x = NULL, z = ~gridmat, opacity = 0.8, colorscale = "Redor", colorbar = list(title = 'Linear fit')) %>%
layout(scene = list(
xaxis = list(
title = '',
tickfont = list(size = 14),
linewidth = 4, # Increase axis line width
gridwidth = 2 # Increase grid line width
),
yaxis = list(
title = '',
tickfont = list(size = 14),
linewidth = 4, # Increase axis line width
gridwidth = 2 # Increase grid line width
),
zaxis = list(
title = '',
tickfont = list(size = 14),
tickvals = c(10, 30, 50, 70), # Specify tick labels
linewidth = 4, # Increase axis line width
gridwidth = 2 # Increase grid line width
)
))
fir## No trace type specified:
## Based on info supplied, a 'scatter3d' trace seems appropriate.
## Read more about this trace type -> https://plotly.com/r/reference/#scatter3d## No scatter3d mode specifed:
## Setting the mode to markers
## Read more about this attribute -> https://plotly.com/r/reference/#scatter-modeFigure 7G-I
Various slices through the 3D scatter plot, for individual TFs
cebpa <- subset(combined, TF1 == "Cebpa")
base_slope <- 100*active_libA$max_active[active_libA$TF == "Cebpa"]
formerge <- active_libA; colnames(formerge)[1] <- "TF2"
cebpa <- merge(formerge,cebpa, by = "TF2", suffixes = c(".alone", ".cebpa"))
qplot(x = 100*max_active.alone, y = 100*max_active.cebpa,data=cebpa) +
geom_abline(intercept = m$coefficients[1] * base_slope, slope = m$coefficients[2] + base_slope *m$coefficients[3] , linetype = 2) +
geom_text_repel(aes(label = TF2), color = "black", data = subset(cebpa, TF2 %in% c("Myb","Sp1", "Mecom", "Nfyc", "Elk1", "Gata1", "Nfkb1", "Spi1", "Stat5a"))) +
theme_bw() + theme(panel.grid = element_blank()) +
xlab("% of sequences active\n(single factor / lib A)") + ylab("% of sequebces active\n(with Cebpa)")
cebpa <- subset(combined, TF1 == "Gata1")
formerge <- active_libA; colnames(formerge)[1] <- "TF2"
cebpa <- merge(formerge,cebpa, by = "TF2", suffixes = c(".alone", ".cebpa"))
base_slope <- 100*active_libA$max_active[active_libA$TF == "Gata1"]
qplot(x = 100*max_active.alone, y = 100*max_active.cebpa,data=cebpa) +
geom_abline(intercept = m$coefficients[1] * base_slope, slope = m$coefficients[2] + base_slope *m$coefficients[3] , linetype = 2) +
geom_text_repel(aes(label = TF2), color = "black", data = subset(cebpa, TF2 %in% c("Cebpa", "Myb", "Nfkb1", "Meis1", "Fli1", "Mecom"))) +
theme_bw() + theme(panel.grid = element_blank()) +
xlab("% of sequences active\n(single factor / lib A)") + ylab("% of sequebces active\n(with Gata1)")
cebpa <- subset(combined, TF1 == "Meis1")
formerge <- active_libA; colnames(formerge)[1] <- "TF2"
cebpa <- merge(formerge,cebpa, by = "TF2", suffixes = c(".alone", ".cebpa"))
base_slope <- 100*active_libA$max_active[active_libA$TF == "Meis1"]
qplot(x = 100*max_active.alone, y = 100*max_active.cebpa,data=cebpa) +
geom_text_repel(aes(label = TF2), color = "black", , data = subset(cebpa, TF2 %in% c("Gata2", "Tcf3", "Tfap2a","Cebpa", "Trp53"))) +
geom_abline(intercept = m$coefficients[1] * base_slope, slope = m$coefficients[2] + base_slope *m$coefficients[3] , linetype = 2) +
theme_bw() + theme(panel.grid = element_blank()) +
xlab("% of sequences active\n(single factor / lib A)") + ylab("% of sequebces active\n(with Meis1)")
print(sessionInfo())## R version 4.4.1 (2024-06-14)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.10
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.12.0
## LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.12.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## time zone: Europe/Madrid
## tzcode source: system (glibc)
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] plotly_4.10.4 ggnewscale_0.5.1 gridExtra_2.3 reshape2_1.4.4
## [5] plyr_1.8.9 ggrepel_0.9.6 ggplot2_3.5.1
##
## loaded via a namespace (and not attached):
## [1] sass_0.4.9 generics_0.1.3 tidyr_1.3.1 stringi_1.8.4
## [5] digest_0.6.37 magrittr_2.0.3 evaluate_1.0.3 grid_4.4.1
## [9] RColorBrewer_1.1-3 fastmap_1.2.0 jsonlite_1.8.9 httr_1.4.7
## [13] purrr_1.0.4 crosstalk_1.2.1 viridisLite_0.4.2 scales_1.3.0
## [17] lazyeval_0.2.2 jquerylib_0.1.4 cli_3.6.4 rlang_1.1.5
## [21] crayon_1.5.3 munsell_0.5.1 withr_3.0.2 cachem_1.1.0
## [25] yaml_2.3.10 tools_4.4.1 dplyr_1.1.4 colorspace_2.1-1
## [29] vctrs_0.6.5 R6_2.6.0 lifecycle_1.0.4 stringr_1.5.1
## [33] htmlwidgets_1.6.4 pkgconfig_2.0.3 pillar_1.10.1 bslib_0.9.0
## [37] gtable_0.3.6 glue_1.8.0 data.table_1.16.4 Rcpp_1.0.14
## [41] xfun_0.50 tibble_3.2.1 tidyselect_1.2.1 rstudioapi_0.17.1
## [45] knitr_1.49 farver_2.1.2 htmltools_0.5.8.1 rmarkdown_2.29
## [49] labeling_0.4.3 compiler_4.4.1